Every day it seems like there are exponentially more articles, videos and podcasts popping up on the topic of AI. Although there may not be a consensus on whether AI is good or bad for the future of the human race, one thing is for certain - AI is not going away any time soon.
Personally, I have to admit that I have been slow to integrate AI tools into my everyday worklife. However, I’m always looking for ways to increase my effectiveness, so I set myself a mission to delve into the world of AI and investigate what opportunities AI tools can offer for Scrum and agile delivery professionals.
This article will focus specifically on the day to day operations of a Scrum Master but will be relevant for most roles & industries. It will outline my findings and provide some practical, actionable tips to support you in implementing AI tools and, as a result, hopefully give you more time to do the things you want to be doing and that add the most value.
The Scrum Master role & responsibilities
The first step in the process was to outline the key responsibilities of a Scrum Master and then identify areas where there might be a place for the use of AI.
According to the Scrum Guide, Scrum Masters are ‘true leaders who serve the Scrum Team and the larger organisation’.
Some common, and often overlapping, pain points that come with the practical realities of serving the Scrum Team and organisation include:
- Limited time due to meetings & other commitments filling diaries
- Important administrative tasks that are time consuming to complete
- Managing workflows
- Progress tracking
- Stakeholder updates
Team communication
Where AI can have the most impact
The next step was to interrogate how AI could help remove or reduce the burden of such pain points. My view had historically been that so much of the Scrum Master role requires a high level of emotional intelligence that I was sceptical about the value of AI in this arena.
However, this overlooked the power of AI in supporting the Scrum Master role in areas where subject matter expertise was not necessary in order to deliver outputs, thus freeing up time for the deep focus work that requires critical and creative thinking.
Some of the key themes that emerged from my research were that AI could be used to:
- Save time on repetitive tasks
- Support decision making processes
- Improve collaboration and communication
Practical tips for harnessing available AI tools
With that in mind, the following section is broken down into a number of areas where AI can be used practically and productively in the completion of both routine tasks, as well as providing insights to support more complex, human-lead activities.
Meetings notes & actions
- Starting with the most straightforward first. There are a number of tools that can provide automated transcripts for calls. At Zoocha we use Google Meet and even with the standard workspace you are able to enable a transcription & recording to be created with a tick of a checkbox in the settings. Other similar tools such as Microsoft Teams offer the same functionality. If Gemini is enabled for your workspace, this now offers a levelled up version where both a transcript and summary is provided - the summary itself can be a little too scarily accurate!
- The transcript provided can also be fed into ChatGPT to provide you with a list of the actions discussed and agreed on the call that you can cross reference with any notes you have.
Emails
- We’ve all been there - an email comes through that is the length of War & Peace and it’s not immediately obvious what the takeaway is. Ask tools like ChatGPT to summarise lengthy emails into concise, digestible summaries.
- Similarly, if you are writing an email and are unsure if it meets the mark, use ChatGPT to review & potentially re-write it, rather than spending hours agonising over the contents. You can even ask it to sense check the tone of voice or adjust the language based on the intended recipient.
Document reviews
- If you have been provided with a new version of a document, you can ask ChatGPT to identify and highlight any key changes between the two versions in order to speed up the review process and potentially avoid missing any changes.
Facilitating collaborative workshops or sessions
- The most common type of session where this is applicable would be the Sprint Retrospective but it would also be useful for any kind of workshop.
- If you use Miro to facilitate your sessions, make the most of the AI features they have rolled out in recent months for sharing the outputs or moving the conversation forward. For example:
- Grouping sticky notes by topic or keyword
- Asking it to summarise key takeaways from the board in a document
- Asking it to suggest ideas for next steps
For the more creative, removing backgrounds from images has provided endless possibilities for new retro templates and the like.
Stakeholder management
- In addition to the suggestions above around email and meeting management, Natural Language Generation (NLG) can be a powerful tool in successful stakeholder management. AI tools can translate complex project data into digestible summaries or presentations - for example provide a detailed sprint report and ask ChatGPT to convert it into a high level summary for stakeholders.
Staying up to date with the latest trends
- It can feel impossible to stay on top of the industry landscape across all the various sources of information available. ChatGPT have rolled out a Beta of ChatGPT tasks, which, amongst other things, can be used to set up weekly updates summarising news from the last week
- An example prompt would be:
- Please track and summarise emerging trends and news in agile software development. Provide a concise weekly summary every Wednesday morning at 8am highlighting key trends and developments. Please include the source of every update.
Routine admin tasks
- How often are you asking a team member whether their Jira ticket is in the right status? If the answer is frequently, then implementing automated workflows into your project management tools with apps like Zapier can help eliminate the need for anyone to be manually updating where something is in the workflow.
- An example of this would be triggering the status of a Jira ticket to change after the linked PR is merged & deployed
- Another common unavoidable time consuming task can be diary management. Tools such as Clockwise can analyse team calendars and propose optimal meeting times for the group, saving you the headache of trying to find a slot for 10 people.
Predictive analytics
Depending on the data you are currently tracking for your sprints & the tools you are using, AI can do the leg work around generating insights and predictive analytics of future trends based on historical performance.
There are different ways to approach this. For example, Jira offers a dedicated product, Jira Insights, that states it is designed to `help the team make data-driven decisions without leaving their current context.` Alternatively feeding key metrics into Gemini or GPT from previous sprints and asking probing questions can similarly provide data that can used for holding up the mirror in retrospectives and dictating sprint & release planning.
Proceed with caution
Of course with any new practices we must not only assess their benefits but also their potential risks in order to mitigate any unforeseen negative consequences. There are a few key areas to highlight when adopting AI tools into your roster:
- A big one is data privacy. Always ensure compliance with GDPR or other relevant data protection regulations when handling sensitive data. For example, ensure that the tools you are using are not learning from your inputs.
- Start small. As with anything new, apply the test & learn approach to any new tools before ramping up your usage.
- Understand any additional costs; some AI extensions or tools require a subscription or an additional cost on an existing subscription, so be sure to review this carefully before use.
- Sustainability; the use of AI tools can be significantly more resource intensive than a Google search or similar. Consider if the output is worth it.
- Trust nothing. I would advise strongly against sharing raw outputs from tools such as ChatGPT without an appropriate review & edit step.
In conclusion, if you are venturing into the world of AI, this post has highlighted that there are multiple ways to ease yourself in and avoid sinking time into repetitive or labour intensive tasks.
As with all things, however, careful consideration must be paid to both the inputs and the outputs to ensure that you are acting in compliance with wider compliance policies and also not affecting the overall quality of your work products.
Most importantly, AI tools on their own are not enough. For example, it can’t tell you why a team’s velocity is reducing but it can flag the early signs to trigger an investigation. AI should be used to enhance and complement human emotional intelligence, judgement and knowledge - not replace it.
Get in Touch
If you’d like to discuss how we use AI or run Agile projects at Zoocha, please feel free to contact us at [email protected].
About the author
As Delivery Director, Hannah ensures standards of quality, accountability and Agile principles are baked into each project delivered by the expansive global team. From mentoring junior Scrum Masters to facilitating the adoption of Agile in large scale public sector institutions, Hannah takes a person-centred approach to opening the doors to Agile and embracing the flexibility of the Scrum Framework.